The staff of the Biostatistics Core will be responsible for providing Biostatistical support to the research of this program. The Biostatistical Core is under the supervision of Dr. Jeremy M.G. Taylor, Director of the University of Michigan Comprehensive Cancer Center Biostatistics Unit. The core provides assistance in the design, analysis and interpretation of preclinical and clinical experiments of the program project. Core personnel will interact with project investigators to ensure that appropriate designs and methods of analysis are used. Design issues involve selection of doses, randomization, timing of measurements, number of animals or patients. For analysis of data the core will ensure that efficient methods are used. Standard graphical, group comparison and correlation methods of analysis will be used for initial investigation of the experimental data. Mixed model methods will be used for efficient use of the data in experiments involving repeated measures. Core personnel are experienced in the design and analysis of both animal and clinical data. This will ensure that all data obtained from MR measurements (Projects 1-4), tumor histology (Projects (1-3), cell kill and growth changes associated with therapy (Projects 1-3) and patient outcome (Project 4) will be collected efficiently and analyzed appropriately.